Form and Function

[Piscean]

1) How do we know what are the possible classes that a word can fall into?

We don't know in advance. When new words come into the language, or existing words are used in different ways, we have to judge how we categorise them - assuming we wish to categorise them.

2) If the word class is at least partly determined by its role in a sentence, then when there is a conflict between function and form, how do we decide on word class when more than one option is available?

If we have a clear idea of our criteria for the categories, then there is normally no conflict. The only conflict is between different schools of grammar. I am sure that for many people, brick in brick wall is an adjective because it clearly modifies wall. For Quirk et al, Huddleston and Pullum, Aarts, and me, it has none of the formal characteristics of an adjective (bricker, brickest, very brick, etc). It is a noun(-modifier).

[3) Is it reasonable to accept that if a word is used in context, in a meaningful sentence, in a way that conflicts with the set of possible classes, then this set of possible classes should be extended to include this?

I am not sure that I understand this. If you are asking if a new word class can be invented if a word is used in a way that cannot be slotted into the existing word classes, then the answer is yes. For our first grammarian, adjectives were a sub-class of nouns. It was not until the end of the 18th century that adjective was universally recognised as a distinct word class. In our lifetimes we have seen the class of determiner/determinative added to the sets of most writers.

When there is a conflict of internal form (as in English), how do we select word class, if not by function?

I think I have dealt with that one.

5) If the class of a word is determined by its form as well as its function, then how much weight should be given to each?

How long is a piece of string?

Writers on grammar do not always agree on which characteristics are key. I, for example, do not go along with H & P on some of the words they have added to their preposition word class. Linguists are not alone in this. Scientists (I am not expert enough in that area to say which type of scientist I should name) disagree on which species or sub-species they should assign to certain living things. I believe that there is some disagreement about whether a virus is a living thing.

One of the interesting, if sometimes frustrating, things for me about grammar is reading the ideas of grammarians on such topics.

[jutfrank]

Okay, I feel we're getting somewhere. Thank you for your responses.

Nothing but the fact that native speakers have so far used act only as a noun and a verb.

Yes. The class is determined by precedented usage. So I wonder how some people could consider up in She put the tent up as a preposition instead of a particle adverb, when there is such heavy and common usage. The same with all other phrasal verb particles, and with good as adverb in response to How are you? and with fast as adverb in drive fast and so on.

Just to be clear about your own position: Are you happy to accept up as adjective in Your computer is up? If so, is this then based on its precedented use as an adjective, ie., the weight of all the times it has ever been used as a modifier? That is, that what determines the word class is usage over form?

Regarding how we can know that good is an adjective:

... from the fact that it has (admittedly highly irregular) comparative forms, and that it can be modified by an adverb.

I would dispute that better and best are forms of good. They both sound completely different and have separate etymologies so it's hard to accept that they are expressions of form of the root adjective good. My claim is that the only measure by which we class good as an adjective is its usage as a modifier.

The difference between good in your two examples is that in Only the good die young, good must imply people or, if we are talking about some other group of living things, that group. In the good of mankind, we understand good as an abstract concept. It is not synonymous with good cause (or any other noun).

Okay. I don't want to take this point any further at the moment because I'm not sufficiently convinced either way. I just wanted to know your position.



To attempt to draw some conclusions, this is the main thrust of my argument so far:

1) The criteria we use when classing words as nouns/adjectives, etc, constitute a range of both formal and functional features. We know the class by what forms it can express and by how it can be used in context.

2) In some cases, there are only functional features which are available to determine word class. Sometimes the only way we can tell the class is by usage. Either a particular use in context or by a general precedented usage.

3) There are always functional features. Language does not exist outside of usage.

4) Therefore, function takes precedence over form as an indicator of word class. When the formal features of a word dictate a certain class, but the use dictates a different class, then usage takes precedence.

Please criticise and agree/disagree.

[jutfrank]

If we have a clear idea of our criteria for the categories, then there is normally no conflict. The only conflict is between different schools of grammar. I am sure that for many people, brick in brick wall is an adjective because it clearly modifies wall. For Quirk et al, Huddleston and Pullum, Aarts, and me, it has none of the formal characteristics of an adjective (bricker, brickest, very brick, etc). It is a noun(-modifier).

Okay, yes, that's clear.

I am not sure that I understand this. If you are asking if a new word class can be invented if a word is used in a way that cannot be slotted into the existing word classes, then the answer is yes.

Yes, that's what I meant.

For our first grammarian, adjectives were a sub-class of nouns. It was not until the end of the 18th century that adjective was universally recognised as a distinct word class. In our lifetimes we have seen the class of determiner/determinative added to the sets of most writers.

I like the determiner/determinative distinction as it separates form and function nicely. If we really want to avoid confusion of form and function, why don't we do this across the board?

How long is a piece of string?

Measure it and find out!

Writers on grammar do not always agree on which characteristics are key. I, for example, do not go along with H & P on some of the words they have added to their preposition word class.

If you don't mind, which words are they? I have a copy of Aarts, which I like, but not an H&P. I would be interested, purely as material for discussion which classes you don't like and why.

By the way, you mentioned a few posts ago Otto Jespersen, for whom I have great admiration. Thank you for alerting me to his The Philosophy of Grammar, which I'd never read but which I've since acquired and started to read. Do you have a copy?

[Piscean]

I like the determiner/determinative distinction as it separates form and function nicely. If we really want to avoid confusion of form and function, why don't we do this across the board?

Linguistics is not a precise science. It can't be - languages generally, and English in particular, are not logically arranged systems. Grammarians work to devise the best possible ways of describing the system, but there can be no infallible way. Until some grammarian comes up with a way that is indubitably better than anything else conceived, we cannot expect universal acceptance of any way existing now.

If you don't mind, which words are they? I have a copy of Aarts, which I like, but not an H&P. I would be interested, purely as material for discussion which classes you don't like and why

.
H & P's expanded class of prepositions includes not only words universally accepted as prepositions but also:

1. the 'subordinating conjunctions' of more traditional grammar such as before and after: He left after you promised to help.
2. prepositions traditionally classed as adverbs when they have no complement: I haven't seen her since.
3. A number of forms traditionally classed as participles: Barring accidents, they should be back today. Given his age, a shorter prison sentence is appropriate.

(Examples from Huddleston and Pullum (2002. Ch7) The Cambridge Grammar of the English Language. Cambridge: CUP)

I am not happy with the first group. I see the force of some of H & P's arguments, but I still believe that such words belong on the conjunction side of the preposition/conjunction border.

By the way, you mentioned a few posts ago Otto Jespersen, for whom I have great admiration. Thank you for alerting me to his The Philosophy of Grammar, which I'd never read but which I've since acquired and started to read. Do you have a copy?

I do indeed. I have hard copies of almost all the books I cite in these forums.

[jutfrank]

H & P's expanded class of prepositions includes not only words universally accepted as prepositions but also:

1. the 'subordinating conjunctions' of more traditional grammar such as before and after: He left after you promised to help.
2. prepositions traditionally classed as adverbs when they have no complement: I haven't seen her since.
3. A number of forms traditionally classed as participles: Barring accidents, they should be back today. Given his age, a shorter prison sentence is appropriate.

This classification makes a lot of sense to me.

1. I have my own basic working definition of preposition: a word which expresses a relationship between things/events. The precise nature of the distinction between things and events I won't go into right now, suffice to say that there is a correspondence to noun phrases (things) and clauses (events). By this view, I would agree that after in the example makes sense as a preposition as it expresses the temporal relation between He left (Event 1) and you promised to help (Event 2). My definition above of prepositions is very much based on the function of the word and not at all on form. I wonder if H & P's reasoning for their classification is similarly based. If so, is there a movement in modern grammar to emphasise function over form as a basis by which to classify words?

2. If I understand correctly, they're suggesting that since is best considered a preposition since there is an implied complement, i.e., 'since the last time I saw her, whenever that was'. Have I got that right? This idea of implied words having a bearing on how we analyse grammar is very interesting to me and one that I, as someone more interested in meaning than form, naturally find appealing. Again, is there a movement in modern to emphasise meaning over form as a basis by which to classify words?

3. It seems to me that H & P may be considering these words as prepositions as they express kinds of logical relation between things. I too would be comfortable accepting them as such, according to my basic definition above of the function of prepositions.

[Piscean]

I wonder if H & P's reasoning for their classification is similarly based. If so, is there a movement in modern grammar to emphasise function over form as a basis by which to classify words?

I haven't kept up with academic journals for some time, so I can't answer that. I do feel, however, that you are worrying too much about function being more important than form, or vice versa. Both are important.

2. If I understand correctly, they're suggesting that since is best considered a preposition since there is an implied complement, i.e., 'since the last time I saw her, whenever that was'. Have I got that right?

I don't think so. What they say (pp 600-601) is: The presence or absence of a complement has no bearing on the classification of the head in [nouns, verbs and adjectives with -(out) complements]. There is no reason to treat [I haven't seen her since the war and I haven't seen her since] differently, and we accordingly take since as a preposition [in both sentences].We also include in the preposition category certain words like downstairs, which never take complements.

This idea of implied words having a bearing on how we analyse grammar is very interesting to me and one that I, as someone more interested in meaning than form, naturally find appealing.

I find the idea of implied words/phrases useful in teaching, but I am extremely wary of it in formal grammar. I believe that we can postulate with confidence only if we discuss the words actually spoken/written. Once we start talking about implication, we are in a way attempting to put ourselves in the speaker's/writer's mind. We have no guarantee that our mind-reading is correct.

Again, is there a movement in modern to emphasise meaning over form as a basis by which to classify words?

I am not sufficiently up-to-date on the literature to be able to answer that confidently.

3. It seems to me that H & P may be considering these words as prepositions as they express kinds of logical relation between things.

I think you need to read H & P yourself. Please don't try to draw conclusions about what they are considering from my brief summaries.

[jutfrank]

I do feel, however, that you are worrying too much about function being more important than form, or vice versa. Both are important.

This is where I perceive we most obviously disagree. For the sake of argument, and of moving the conversation forward, let me put my view forward in its very strongest form.

Language is fundamentally about meaning. Meaning comes from use. Language can only have meaning as it is used by people in context. The learners on this forum have to learn how to do things with words, in order to have the desired effect on others. For production, they must learn how to use language to say and write what they mean. For reception, they must also learn how to infer meaning from the use of others. The way we use language is what is paramount.

Form is wholly arbitrary. That's why we can communicate the same ideas in French, Urdu, Japanese and Navajo, using completely different sound patterns and scripts. Sure, you have to learn how to pronounce and spell words accurately, and learn how to comprehend and read, but that is only a very small part of learning how to be a competent and effective user, I would argue. Learners don't need to know whether a word is an adjective—they need to know that it's modifying something.

Put simply, grammar is the way that words combine to create structure, which in turn gives meaning. As an academic pursuit, it's the study of how words function to produce meaning. Word form is culturally determined and therefore ephemeral by nature, constantly changing. Grammatical structure is immeasurably more consistent across individual languages and across time. The form that any language takes is just a curious by-product of cultural differentiation, somewhat interesting to a minority of linguists, but ultimately of little importance compared to structure.

[Piscean]

I don't see any strong disagreement between our views in the first two paragraphs of post 17.

Form is wholly arbitrary. That's why we can communicate the same ideas in French, Urdu, Japanese and Navajo, using completely different sound patterns and scripts.

That is not entirely true. For possibly cultural reasons, it is sometimes quite difficult to express precisely an idea conceived in the mind of a speaker of one language in a completely comprehensible way in another language. However, I don't think this is relevant enough for us to examine.

Word form is culturally determined and therefore ephemeral by nature, constantly changing. Grammatical structure is more or less fixed.

Those are rather sweeping statements. I don't think I would accept either without some expansion.

The form that any language takes is just a curious by-product of cultural differentiation, somewhat interesting to a minority of linguists, but ultimately of little importance compared to structure.

I don't see what you are getting at there. I don't I don't know what difference you feel between form and structure.

In any case, I thought we were discussing the importance of form and function in establishing word classes, i.e., to oversimplify, why is happy an adjective? Is it because it has comparative and superlative forms (form) or because it modifies nouns (function)? I maintain that both form and function are important criteria in establishing the class of a word.

[Tdol]

Isn't meaning akin to a marriage of structure and word? I don't see it as the sole generator of meaning.

[jutfrank]

In any case, I thought we were discussing the importance of form and function in establishing word classes, i.e., to oversimplify, why is happy an adjective? Is it because it has comparative and superlative forms (form) or because it modifies nouns (function)? I maintain that both form and function are important criteria in establishing the class of a word.

Okay. I don't know if I might be confusing terminology here.

So you're saying that word class (such as 'adjective') is based on both form and function. So then 'word class' is not a label that describes form, right? This is what I can't understand. If both form and function have a bearing on class, then any class, such as 'adjective' cannot be a label of form.

I think up until now I've been considering word class as a classification of form. So that 'determiner' is a word class but 'determinative' and 'modifier', since they relate only to function, are not.

Before I go on, do you follow?